August 12, 2020
12 min read
Mixing audio for a venue is a well-documented practice. At its most basic level, it requires microphones, prerecorded audio sources, a soundboard, loudspeakers, and, most importantly, a skilled operator. The two most common challenges with live-venue audio are avoiding and eliminating feedback and ground-loop hum.
Mixing audio for a video recording or live stream involves many of the same techniques and equipment, but each has different production standards and you don't have to worry with live sound reinforcement using loudspeakers. The two most common challenges with audio for a live stream are attaining a clean signal and mixing audio within a smaller range than you might if you're only mixing for the venue.
Neither of these two is very complicated to explain and implement, although, in practice, mixing live audio can be very challenging because of room dynamics, electrical grounding issues, and, most importantly, microphone technique.
As a live-stream specialist, my perspective is going to be different compared to how a live audio technician would approach venue audio. The main reason is that I'm considering both the in-person audience and the live-stream audience equally and simultaneously.
In this article, I explain my approach to mixing audio for both of these environments at the same time. In my business, I refer to events that have both an in-person audience and a live online audience as a hybrid event, and the majority of the work that we produce is for hybrid events with dual audiences. Over time, we've found that we prefer to handle the audio for both audiences, rather than trust part or all of this to an external A/V company. The biggest reason is that the needs of the audience in the room are different from the needs of an online viewing audience, and if an audio technician is mixing audio only for the room and is not involved in or monitoring the audio for the live stream, the audio levels may not be within a desirable range.
So, with that lengthy preamble out of the way, let's start with the basics of mixing audio for a live venue.
Step 1: Connecting audio sources
Connecting microphones and prerecorded audio sources to a soundboard is best done with balanced connections like XLR or ¼" TRS cables. The use of balanced connections helps reduce external noise from electromagnetic interference, which is especially an issue with longer cable runs.
Step 2: Setting the pre-amp levels
When you connect a cable to your soundboard, regardless of whether it's a digital board or an analog one, you need to set the preamplifier levels for each input. The pre-amp levels can also be labeled as gain or trim. Some boards have a levels indicator that is green when you're in the correct range and red if your signal is too hot. With all audio signals, you never want to clip your audio anywhere in your signal flow. This is similar to protecting your highlight when shooting video.
Step 3: Confirming clean audio
You can confirm clean audio with headphones or speakers. This is when you listen for obvious problems like a ground loop hum or a noisy or distorted audio signal. Ground loop hum can occur when you're connecting A/C-powered audio devices together. Normally, equipment grounds to the A/C plug it's directly connected to, but sometimes, due to differences in grounding potential, the device chooses the path of least resistance through your audio cables, not to the ground on the plug. This causes ground loop hum.
This electrical interference hum needs to be eliminated. The solution is to add a direct box into the workflow to lift the ground and force the devices to ground, as intended, to their respective plugs. Using a DI should be done automatically when you are connecting audio from a remote laptop both to lift the ground and to convert the unbalanced signal to a balanced XLR connection.
Step 4: Setting PA speaker and soundboard volume levels
I start with my PA speakers set to onequarter volume and then I increase the volume levels for each input and the main output. If your levels are set properly, you should be operating in a range where the levels for the inputs and output are just below the parity level of 0 on the level faders. If your fader levels are too high, then you don't have any ability to increase the levels as needed, and if they have to be set too low, this means that your loudspeakers are set too high and the signal for the live stream won't be within a desirable range.
It might take a bit of back-and-forth to find the right balance between the levels on your loudspeakers, input volume, and main output volume, but this is all a part of mixing professional audio. In this step, you are generally getting the audio into a proper range so that you can move on to the next step; you can then return to this step to get your final show levels set.
Step 5: Equalizing your audio
Generally, this step is part of Step 4 because, most of the time, you will find that when setting your audio levels, you get feedback before the audio gets to show levels. This is especially relevant when you have multiple microphones and especially with omnidirectional microphones like lavaliere mics. Feedback is what happens when certain audio frequencies bounce around the room and back into the microphones at a higher level than the other frequencies. This creates a feedback loop that instantly turns into a mounting squeal. It has to be terminated immediately by manually turning down the levels.
Every room is different in terms of which frequencies will bounce around more than others. Some audio technicians claim they can hear and identify the feedback frequencies, but I prefer to rely on simple frequency spectrum analyzers to identify exactly which frequencies are ringing. If I am using a digital soundboard like the Behringer X18, I can consult the 100-band Real Time Analyzer (RTA) to monitor and identify the problem frequencies, then use the equalizer to turn down the levels only on those specific frequencies. This is more precise than turning down an entire range of frequencies, like you do on most analog boards, and is one of the reasons I prefer digital soundboards.
Audio equalization (EQ) can be done on individual inputs or the main output as a whole. Normally, you would add a bit more EQ on those previously mentioned omnidirectional microphones that are more prone to feedback and less on a cardioid handheld microphone. You also want to be careful not to go overboard with carving out too much when you are EQ'ing, or the resulting signal may sound too much like it was recorded in a fishbowl.
On an analog soundboard, you'll be more limited with how surgically precise you can get. Most boards limit you to only three controls: high, medium, and low. I find this too restrictive, but this is what most smallvenue operators use, and it reinforces my preference to manage my own audio.
My favorite line of analog mixers is the Mackie ProFX V2 line because the eight- and 12channel models that I use have a seven-band stereo graphic equalizer, and this allows me to be more precise than with a traditional three-channel equalizer. Unfortunately, the new V3 line eliminated the graphic equalizer in favor of a compressor.
Before I moved to a digital mixer with a 100band graphic equalizer, I used the Behringer FBQ1502HD 15-band EQ. Although it does feature feedback detection that lights up the frequencies that have the strongest signals, I found that there were too many false positives, and I got better results using an external spectrum analyzer. This might sound like a fancy piece of tech, similar to an expensive light meter for exposure, but I'm talking here about a free, downloadable cellphone app called Advanced Spectrum Analyzer PRO for Android that is also ad-free.
EQ'ing a room is as simple as slowly increasing the volume on your soundboard until you start to hear feedback. Then, using the RTA or a Spectrum Analyzer, you determine which frequency is spiking and causing feedback. Next, on your graphic equalizer, lower that specific frequency. You can then increase the volume again to target the next-worst frequency.
Live-Stream Audio Considerations
Up until this point, I have only been focused on the audio for the venue. Now is the time to start considering audio for the live stream. If you are controlling the soundboard, you can utilize one of your soundboard outputs for your webcast audio. Normally, I prefer to have some control over the levels of each signal. In the venue, you have a much wider acceptable audio range than you do on a recording. The reason is that your PA speakers are much more powerful than laptop speakers, and venue show levels are generally louder than what a typical laptop can reproduce.
This gives you a much wider volume range before your audience struggles to hear what the presenter is saying. With audio on a webcast, you have to assume that the viewer's ability to turn up their speakers is very limited, so if you want the viewer to be able to hear properly, your audio needs to be mixed within a much smaller range. Mackie had the right idea when it added a compressor to its V3 line, understanding that this was desirable for the growing field of digital content creators. However, for my workflows, I'll take the graphic equalizer on the board, knowing I can add a compressor or a more advanced audio plugin later in my digital workflow.
I'm now going to cover three common options for getting analog soundboard audio into your live stream.
Music and movies benefit from stereo and surroundsound mixing, but most of the time, you're going to send the same mono audio signal to both the left and right speakers when you are reinforcing microphone audio for conferences and keynote speeches. As a result, you might daisychain your signal from one speaker to the next, leaving you with a second, unused main output. Or, your soundboard could have a second set of main outputs, typically ¼" TRS outs, as opposed to the XLR mains.
If you are taking a main output signal, you need to realize that both your house and livestream signals are tied together. You then want to capture that audio in a device that gives you sufficient control over your audio levels for the webcast, which will change as you adjust the audio levels for the house. This might mean a second soundboard, but it could also include an external USB audio capture device, a video switcher, or even simply your video camera.
If you want to avoid having your audio tied to the house audio, you can use an Aux output. This could work for jobs with few audio sources, and it normally means if you need to adjust levels on an input, you would need to perform this task twice: once for the live stream and again for the house. Although, as I have mentioned before, since the house levels are more forgiving, you may not need to adjust them as often as you do for the live stream.
Some audio technicians are figuring out that if they set up their soundboards "backwards," they have better control over the livestream levels and don't impact the house levels. Instead of using the main outs for the PA speakers, they are using the mains for the livestream signal and are using the Aux for the house. The reasoning behind this is that the VU meters show the main levels, and having a visual representation of the levels for the livestream levels is important because the technicians can better monitor and adjust those levels with a calibrated meter. They don't need a VU meter for the house levels because they are in the room and can adjust levels by what they are hearing.
If my soundboard is right next to my webcast laptop, then I typically will use the USB output on my analog or digital soundboards. On my Mackie analog board, the USB levels are tied to the input volume levels, not to the main volume levels, so I can still ride the input faders to adjust both mixes at the same time. I also still have the ability to adjust the signal to the loud speakers with the main up or down.
I can't always count on my soundboard being next to my laptop. One of the benefits of having a digital soundboard is that you can control it using a tablet, laptop, or control surface. For lectern microphones and panel discussions, I often use multiple wired gooseneck microphones. While I do have large audio snakes, or I could run multiple lines of XLR cable, it is a lot easier to use the soundboard as a stage box and have it closer to the microphones and their wired connections. I then can use Wi-Fi or run a single line of Ethernet back to my location and operate the controls using software or a control surface.
Secondary Audio Capture Devices
If you don't have access to (or your main soundboard doesn't have) a USB output, you have a few options for getting an XLR audio feed into your live stream. If you're lucky enough to have a video switcher with XLR inputs and proper level controls, this might be the audio solution for you. The two switchers that I use for this are the Roland V-60HD and the Blackmagic Television Studio Pro 4K.
By embedding your audio signal into your video program feed, you don't need a dedicated audio capture device to capture audio into your live stream. Embedded audio will likely be the best workflow if you're using a hardware encoding appliance like the AJA HELO or a Matrox Monarch.
I'm going to quickly mention that while you can also use your video camera's XLR input as a way to embed the audio into your HDMI or HDSDI signal, I prefer that these controls and the responsibility to monitor and adjust these levels lie with my audio or webcast technician, not solely with my camera operator.
USB Audio Devices
If you're using software like vMix or OBS to mix, record, and webcast, you will need a video capture card with embedded audio or a dedicated audio capture device if your audio has not yet been embedded. One common hardware device for capturing XLR audio to a connected webcast computer is the Behringer UMC line of USB audio interfaces. The line has models with one, two, and four XLR inputs, allowing you to mix multiple channels of audio in your webcast software.
Recently, I misplaced my Behringer UMC404HD, and when I was looking into replacing it with another one, I came across the Roland Rubix line. This line is more expensive than similar Behringer models, but I fell in love with some of the features and had to buy one. Ultimately, I found my UMC404HD. It was in the same properly labeled case where I left it, but once I found a better tool for the job, I just had to upgrade. Let me explain why.
The Rubix line's first standout feature is the green/red LED lights for each input and the output. These LEDs give a visual indicator that you have a signal and that it isn't clipping. The second feature is the compressor/limiter control. On my Rubix44, which has four XLR/¼" combo inputs, there are two compressor/limiter controls-one for the first pair and another for the second pair. The front knob allows me to set the threshold. On the rear panel, there are three settings for this control that allow me to select two different compressor settings and one compressor with a limiter setting.
Having a hardware compressor and limiter for my soundboard feed means that I can get strong audio levels without the risk of clipping audio. Without a compressor and limiter, you have to be more cautious. But whereas broadcasters use -12dB for their average peak audio level, webcast producers need to be more aggressive, or their viewers on weaker devices will complain that it isn't loud enough.
The third feature that the Roland Rubix has is a ground/lift switch. There is nothing worse than a ground loop hum entering your audio feed. This problem is easily and traditionally solved with a dedicated DI for each input, but having this function built into my audio capture device is one less piece of gear that I need to worry about and helps to justify the additional cost over the more popular Behringer UMC line.
Mixing audio for a panel discussion with multiple open microphones can get really tricky. Overall, there is too much room noise from all of the microphones, and you're more prone to feedback. Mixing on an analog soundboard means you are likely riding the inputs constantly. Fortunately, there is a better way. My Behringer X18 digital mixer has an auto-mixer inspired by the patented work of Dan Dugan. (Fortunately for me but not for Dugan, his original patents have expired, and companies like Behringer, and even Roland on the V-60HD video switcher, are free to use this technology.)
Auto-mixing automatically gates each audio channel and activates only the ones being used. The attack speed to open the channel when someone speaks is faster than how quickly an operator can react and results in no detectible clipping. It is also able to determine which is the primary microphone. If the two microphones next to it are picking up weaker signals from the same speaker, it is smart enough to clamp them down too. The result is better, cleaner, and stress-free audio that is easy to implement.
Even if you have full control over the microphones, soundboard, and features like compressors and auto-mixing, you might still benefit from adding a few plugins into your workflow. But if you don't have full control, these plugins can be a lifesaver. Here are a few of the Waves audio plugins that I purchased and use in vMix to improve my audio.
NS1 Noise Suppressor or WNS Noise Suppressor
Both of these are similar and are able to differentiate between vocals and noise and then allow you to eliminate the noise. Just be careful not to get too aggressive; this is because some noise sounds more natural than too clean of a signal, and going too far will result in an underwater sound.
This can be used in addition to or instead of a compressor/limiter. You set your desired levels, and the vocal rider automatically rides the volume to keep your audio levels more consistent. This is especially helpful when you have multiple audio sources to juggle. I tend to pair this plugin with a noise suppressor to avoid pumping the noise when someone isn't speaking.
Video producers often overlook the importance of audio in their video productions, but it is a well-accepted fact that bad audio is worse than bad video. When managing hybrid events, you have to consider both the house audio and the live-stream audio with equal consideration. Regardless of whether your workflow uses an analog or digital soundboard, getting familiar with and using tools like the graphic equalizer and compressor/limiter and plugins like an automizer, vocal rider, and noise suppressor will help you deliver clean and strong audio to both your in-room and online audiences.
Shawn Lam is an award-winning video producer and technical director. His Vancouver-based video production company, Shawn Lam Video, Inc., specializes in corporate and event video production, including online video, video switching, webcasting, and video SEO.